* An earlier version of this paper, entitled “Ancient Susu: A Reassessment of Evidence and Identification,” was presented at the meeting of the African Studies Association in Seattle, November 1992. My thanks to David Conrad and Mamadou Diawara for comments on early drafts and to David Henige for giving this paper its editorial finishing. As all other historians concerned with the Western Sudan and with Upper Guinea, I am indebted to P.E.H. Hair, Nehemia Levtzion, Avelino Teixeira da Mota, and other scholars, who labored to make historical sources easily accessible by publishing critical editions and translations.

1. I employ the name “Mali” of the Arab chroniclers, imparted to them by Soninke informants, for the empire called “Mande(ng)” in oral tradition, to distinguish it from the core “Manding” area, and in this paper I use the historical ethnonym “Mandinga” for Mali's ethnic core, including subgroups such as the Senegambian Mandinka and the Bambara.

2. Publications locating Susu in Kaniaga include Nehemia Levtzion's classic Ancient Ghana and Mali (London, 1973), 51.Google Scholar It incorporated the Susu ethnic group into this hypothesis by arguing that “the Soso were forced to leave their old territory,” Kaniaga, “and migrated en masse” to their present habitat near the coast.

3. Vigué, M., “Les peuplades des rivières du sud de la Sénégambie et les erreurs des ethnographes,” Revue scientifique, no. 15 (1888), 456.Google Scholar Arcin claimed a linkage between Jalo/Concho and the ruler of oral tradition's Susu: Arcin, André, Histoire de la Guinée Française (Paris, 1911), 70Google Scholar; cf. my section VI. Famechon wrote, enigmatically, about the Jalonke of the Solimana/Firiya area, “qui appartenaient à un grand état occupant les deux rives du haut et moyen Niger vers le XIV siècle:” Famechon, Lucien-Marie-François, Notice sur la Guinée Française (Paris, 1900), 51.Google Scholar

4. For example in Levtzion, Ancient Ghana, 228n51, and in Niane, Djibril Tamsir, “Recherches sur l'empire du Mali au moyen age,” Recherches africaines (1960), 23.Google Scholar

5. I consulted versions of the Sunjata epic published in Quiquandon, F., “Histoire de la puissance mandingue,” Bulletin de la Société de géographie commerciale de Bordeaux (1892), 305–429Google Scholar; de Zeltner, Franz, Contes du Sénégal et du Niger (Paris, 1913)Google Scholar; Vidal, J., “La légende officielle de Soundiata,” Bulletin du Comité d'études historiques et scientifiques de l'Afrique Occidentale Française 7 (1924), 317–28Google Scholar; Frobenius, Leo, Dichten und Denken im Sudan (Jena, 1925), 303–43Google Scholar; Monteil, Charles, “Les empires du Mali,” Bulletin du Comité d'études historiques et scientifiques de l'Afrique Occidentale Française, 12 (1929), 291–447Google Scholar; Sidibé, Mamby, “Soundiata Keita, héros historique et légendaire, empereur du Manding,” Notes africaines, 82 (1959), 41–50Google Scholar; Niane, D. T., Sundiata, an Epic of Old Mali (London, 1965)Google Scholar, a translation of Soundiata, ou l'épopée Mandingue (Paris, 1960)Google Scholar; Kanoute, Dembo, Histoire de l'Afrique authentique, trans. Sanogho, T. and Diallo, I. (vol. 1: Dakar, 1972)Google Scholar; Innes, Gordon, Sunjata: Three Mandinka Versions (London, 1974)Google Scholar; Condé, Alpha, Les sociétés traditionnelles mandingues (Niamey, 1974)Google Scholar; Camara, Laye, Le mâitre de la parole (Paris, 1978)Google Scholar; Johnson, John, The Epic of Son-Jara. A West African Tradition (Bloomington, 1986)Google Scholar; Cissé, Youssouf Tata and Kamissoko, Wa, La grande geste du Mali. Des origines à la fondation de l'Empire (Paris, 1988).Google Scholar

6. In some versions of the Sunjata epic the hero is called “Mari Jata.” I attribute this to “feedback,” the contamination of a source by the incorporation of elements from other sources. Niane, , Sundiata, 35–36Google Scholar; Sidibé, ,“Soundiata,” 41Google Scholar; and Camara, Maître, 130 et passim, occasionally substituted “Soundiata” with “Mari Jata,” probably because they knew of Delafosse's or Monteil's identification of Sunjata with Mari Jata. In his composite history Tautin divided the hero into two: Mari Jata chasing the Soso and Sunjata killing Sumanguru, thus tacitly reconciling the different names found in Ibn Khaldun's chronicle and in the oral traditions he himself had recorded in 1887: Tautin, L., “Légende et traditions des Soninké relatives à l'Empire de Ghanata,” Bulletin de géographie historique et descriptive, 9/10 (1894/1895), 476.Google Scholar

7. de Barros, M. M., “Guiné Portugueza ou breve noticia sobre alguns dos seus usos, costumes, linguas e origens de seus povos,” Boletim da Sociedade de Geographia de Lisboa 12 (1882), 720.Google Scholar

8. Quiquandon, , “Histoire,” 309.Google Scholar

9. Delafosse, Maurice, Haut-Sénégal-Niger (3 vols.: Paris, 1912), 2:163–71.Google Scholar

10. Tautin, , “Légende,” 477n1Google Scholar; Gallieni, Joseph, Deux Campagnes au Soudan français (Paris, 1891), 592Google Scholar; Mage, Eugène, Voyage dans le Soudan occidental (Sénégambie-Niger) (Paris, 1868), 161 (“Soso ou Masoso”)Google Scholar

11. Thus, Niane's informant calls the king of Do “Faony Kondé, Faony Diarra” (Niane, , Sundiata, 55Google Scholar). Correspondences of clan names possibly derive from alliances between powerful lineages, which were subsequently extended to the clan level, and from the need for travelers (traders, etc.) to establish pseudo-kin relations in areas without ‘relatives’ of their own clan.

12. Delafosse, , Haut-Sénégal-Niger, 2:163–65.Google Scholar A similar association of Sumanguru with the Jariso is claimed in a version of the Sunjata epic collected in 1963 (Camara, Maître, 89, 178): a case of feedback or evidence that Delafosse only transmitted a view held by some griots?

13. Conrad, David, “Oral Sources on Links Between Great States: Sumanguru, Servile Lineages, the Jariso, and Kaniaga,” HA 11 (1984), 41–42, 45.Google Scholar

14. Monteil, , “Empires,” 354Google Scholar; similar identification in Vidal, “Légende,” 321.

15. Location of Susu in Frobenius, , Dichten, 332, 336, 338Google Scholar; and the Wagadu epic in Frobenius, Leo, Spielmannsgeschichten der Sahel (Jena, 1921), 49–162.CrossRefGoogle Scholar

16. Monteil, , “Empires,” 354–55.Google Scholar For more references and a discussion see Conrad, , “Oral Sources,” 39–41.Google Scholar See also Meillassoux, Claude, Doucoure, Lassana, and Simagha, Diaowé, Légende de la dispersion des Kusa (Epopée Sonike) (Dakar, 1967).Google Scholar

17. Ibid.

18. David Conrad, personal communication.

19. Corpus of Early Arabic Sources for West African History, ed. and trans. Levtzion, N. and Hopkins, J.F.P. (Cambridge, 1981), 477Google Scholar: “One learned in fiqh,” the “practical rules derived from the Shari'a.”

20. Ibid., 333, 334.

21. Ibid., 322.

22. Ibid., 334. In those days “Takruri” denoted persons from the Western Sudan.

23. Ibid., 333.

24. Ibid., 345-46, 355.

25. Ibid., 261. Al-Qalqashandi added Susu to this list, which he must have assumed to be a province of Mali, probably following the Mari Jata tradition recorded by Ibn Khaldun (ibid., 345).

26. Ibid., 322.

27. Ibid., 333.

28. Ibid. Often Arab and Portuguese authors did not distinguish between the name of a kingdom and that of a people. They employed “Susu” both as a region's name and as an ethnonym.

29. Al-'Umari in ibid., 262.

30. The Malian expansion into the region adjoining the ocean is featured in the tradition recorded by Ibn Khaldun, as is the Malian province of Jolof in a regional tradition recorded in ca. 1615 (see section VI), and in recent traditions the victorious campaign of one of Sunjata's military leaders, Tiramakhan Traore, against Jolof. Also, a Malian king prided himself on the vastness of their realm and that it was “contiguous with the Ocean” (al-'Umari in ibid., 267). He made a point of mentioning a naval expedition into the Atlantic Ocean organized by his predecessor (ibid., 268). The mid-seventeenth century tradition of Songhay also named the environment of “Baghena” (Ghana) “as far as the Ocean” among the conquests of Mali: EsSa'di, , Tarikh es-Soudan, trans. Houdas, O. (Paris, 1900), 20.Google Scholar

31. For the transformation from panegyric to epic see Ralph Austen, “The Historical Transformation of Genres: Sunjata as Panegyric, Epic(s) and Novel,” paper presented at the Sunjata Conference, Northwestern University, Evanston, November 1992. In the process much of the tradition's celebrating royal rituals and genealogy was omitted, while heroic episodes and elements reflecting popular beliefs, values, and ideas were included.

32. See ibid. David Conrad, “A Town Called Dakajalan: The Sunjata Tradition and the Question of Ancient Mali's Capital,” 6, and Stephen Bulman, “An Examination of the Role of the Literary Mediator in the Dissemination of the Sunjata Epic,” papers presented at the same conference. I use the concise term ‘pagan’ without any disparaging connotation, other terms being unwieldy or incorrect.

33. Corpus, 322, 421n1.

34. Lange, Dierk, “Das alte Mali und Ghana. Der Beitrag der Oraltraditionen zur Kritik einer historiographischen Fiktion,” Historische Zeitschrift 255 (1992), 596, 598.Google Scholar

35. For example in Corpus, 333.

36. Ibid., 322.

37. Triaud concludes from his interpretation of al-Bakri that Muslim merchants did not frequent the Bure gold fields in the mid-eleventh century. Hunwick agrees: Hunwick, J.O., Meillassoux, C., and Triaud, J.-L., “La géographie du Soudan d'après al-Bakri: trois lectures” in Le sol, la parole et l'ecrit. Mélanges en hommage à Raymond Mauny, ed. Devisse, Jean (2 vols.: Paris, 1981) 1:410, 418.Google Scholar Al-Sharishi, who died in 1222, had information from merchants who had been in Ghana, which they termed the “furthest point reached by merchants” (Corpus, 152). In 1337/38 al-'Umari reported that an informant had lived in the capital of Mali “for 35 years and went to and fro in this kingdom” (Corpus, 262), i.e., since ca. 1300. It must thus have been in the course of the thirteenth century, probably with the rise of Mali, that Arab merchants began traveling south of the early entrepôts of the gold trade in the Upper Senegal area.

38. Corpus, 322

39. Ibid., 332-33

40. Hunwick, , “Géographie du Soudan,” 420.Google Scholar

41. With the exception of Wa Kamissoko (according to Y.T. Cissé in his Avant-propos to Cissé/Kamissoko, , Grande Geste, 27)Google Scholar, possibly a case of ‘feedback.’

42. Sceptics have argued that the constant process of recreating oral traditions left nothing unchanged and that thus nothing in it is to be treated as historical evidence (Johnson, , Son-Jara, 63–64n4).Google Scholar The survival of the Mali-Susu conflict in a tradition since at least the late fourteenth century shows that the essence of certain core elements may, despite all changes, be preserved over long periods.

43. For a structuralist interpretation of elements of the Sunjata epic see Jackson, Michael, “Prevented Successions: A Commentary Upon a Kuranko Narrative” in Fantasy and Symbol, ed. Hook, R.H. (London, 1979), 95–131.Google Scholar

44. Johnson, , Son-Jara, 46Google Scholar

45. Cf. the editors' “Introduction” in Barber, Karin and Farias, Paulo F. de Moraes, eds., Discourse and Its Disguises: The Interpretation of African Oral Texts (Birmingham, 1989), 2.Google Scholar

46. For example in Cissé/Kamissoko, Grande Geste, 154/157, and Niane, , Sundiata, 38.Google Scholar His image in the griots' narrations is also that of the archetype of a griot, as the owner of a magical balafon, and of a leatherworker, owning hides and shoes made from human skins (for instance, in Niane, , Sundiata, 39Google Scholar), but these images seem very secondary.

47. Doumbia, Paul-Emile-Namoussa, “Etude du clan des forgerons,” Bulletin du Comité d'études historiques et scientifiques de l'Afrique Occidental Française, 29 (1936), 340.Google Scholar

48. For a different interpretation of the similar blacksmith image of Fakoli in the Sunjata epic see Conrad, David, “Searching for History in the Sunjata Epic: The Case of Fakoli,” HA 19 (1992), 180–86.Google Scholar

49. Shifts between w and g, k, and j are a common linguistic feature in the wider region: cf. the variants Gangara and Wangara (a region and an early name for merchants in long-distance trade), Jolof and Wolof (a region and a people in northern Senegambia), kuru and woro, “kola nut” in West African languages.

50. In the region clans are non-corporate, exogamous groups of lineages sharing a name, a totem, and, so the members believe, a common ancestor.

51. Delafosse, Maurice, La langue mandingue et ses dialectes (Malinké, Bambara, Dioula) (2 vols.: Paris, 1929–1955), 2:241.Google Scholar I suspect that al-Dakri's “Samaqanda” (Corpus, 81) was the kanda of Sama and al-Bakri's king of Gao “called Qanda” (Corpus, 87) also referred to the title, although Lange tends to believe that it was the name of the pre-Almoravid dynasty (Lange, D., “Les rois de Gao-Sané et les Almoravides,” JAH 32 1991, 269–70CrossRefGoogle Scholar). In Senegambian Mandinka kanda denotes a person of influence based on charisma and/or wealth, not on official rank, and ngana denotes the military leader or hero.

52. Innes, Sunjata, 239-40, 271, 311, 147; Frobenius, , Dichten, 332, 340.Google Scholar

53. Other clan names derived from titles: Masari (“member of the royal lineage,” an alternative of Keita, cf. Senegambian Manka(li)/Masali); Tunkara (Soninke tunka “king”); Samura (royal lineage of Solimana, possibly suma “king” plus suffix -ra, cf. Tunkara); Jagara (“member of the royal lineage,” Papel of Bissau).

54. Segou and Keleyadugu: Jean-Loup Amselle and Jean Bazin, “Présentation;” Jean Bazin, “Princes désarmés, corps dangereux. Les ‘rois-femmes’ de la région de Segu;” Amselle, Jean-Loup, “Un état contre l'état: le Keleyadugu,” all three in Cahiers d'études africaines, 28 (1988), 326, 386n28, 466CrossRefGoogle Scholar

55. Gomes, Diogo, De la première découverte: récit par Diogo Gomes, ed. Monod, T., Mauny, R., and Duval, G. (Bissau, 1959), 37/39 (f. 277).Google Scholar Besides, “Semanagu” sounds very much like “Somanogo,” a variant of “Sumanguru”—evidence that this was a current name?

56. Frobenius, , Spielmannsgeschichten, 17.Google Scholar Possibly the ordinal first name Samba for the first or second son and the southern Senegambian clan name Sambu are derived from the title.

57. Es-Sa'di, , Tarikh es-Soudan, 20, 155, 278.Google Scholar Judging from the context, the “Sanqari-Zouma” seems not to have been associated with the Fulas of “Sanqar” north of Jenne, their ruler “Sanqara-Koi” or his land “Sanqara,” mentioned in es-Sa'di's account (ibid., 248-49, 283-85, 297). In identifying the latter part of the title as suma, I disagree with Heinrich Barth, who considered the final syllable -ma to be a suffix: Barth, , Reisen und Entdeckungen in Nord- und Centralafrika in den Jahren 1849 bis 1855 (5 vols.: Gotha, 1857–1859), 4: 612n.Google Scholar

58. Johnson, John, “The Epic of Sun-Jata According to Magan Sisókó” (PhD., Indiana University, 1979), 2:145Google Scholar; Moser, R. E., “Foregrounding in the Sunjata, the Mande Epic” (PhD., Indiana University, 1974), 279.Google Scholar Both references reported by Wilks, Ivor, “The History of the Sunjata Epic: A Review of the Evidence,” paper presented at the Sunjata Conference, Northwestern University, November 1992.Google Scholar

59. Kanoute, , Histoire, 31Google Scholar, (“Nalo Naré” = Nalu?). In this tradition the Baga are also linked to the Futa Jalon. Traditions of the litoral Nalu, Landuma, and Baga have them as originating from Futa Jalon. I believe that this must be interpreted as an ‘adopted origin.’ The motif of migration from Futa Jalon reflects the wish to borrow the historical prestige of the Futa Jalon, first as part of the Susu empire, later as the site of the Muslim Fula revolution.

60. Sumanguru in the south: Frobenius, , Dichten, 329Google Scholar; Susu west of Niani-Ba: ibid., 332, 336, 338; the exception, Susu east of “Njani Mba” ibid., 340.

61. Johnson, , Son-Jara, 218Google Scholar

62. Kanoute, , Histoire, 36Google Scholar; Suso, Bamba and Kanute, Banna in Innes, , Sunjata, 79, 209, 211, 271, 273Google Scholar

63. Person, Yves, Samori: une révolution Dyula (3 vols.: Dakar, 1968–1975), 1:72, 350–51Google Scholarn33.

64. Ibid., 1:348n14

65. Humblot, P., “Du nom propre et des appellations chez les Malinké des vallées du Niandan et du Milo,” Bulletin du Comité d'études historiques et scientifiques de l'Afrique Occidental Française, (1918), 531Google Scholar

66. Intermarriage between Konte and Konate and hence separate clans: Humblot, , “Nom,” 529, 531Google Scholar; interchange of names Konte and Konaté for Sankaran kings' ancestor: Zeltner, , Contes, 41, 42Google Scholar

67. Caillié, René, Journal d'un voyage à Temboctou et à Jenné (3 vols.: Paris, 1830), 1:335–36Google Scholar; Laing, Alexander Gordon, Travels in the Timannee, Kooranko, and Soolima Countries in Western Africa (London, 1825), 371Google Scholar

68. Person, , Samori, 1:72Google Scholar

69. Camara, Maître, 37n3

70. Yves Person is one of the few exceptions. But while Samori, his grand monograph on the late precolonial period, was the result of meticulously collecting and using all available types of evidence, his notion of early history was one of migrations of ethnic groups and clans taken from selected oral traditions. See, for example, Person, Y., “Les Kissi et leurs statuettes de pierre dans le cadre de l'histoire ouest-africaine,” BIFAN 23B (1961), 1–59.Google Scholar The only writing witness known to have visited Sankaran before the colonial conquest was Winwood Reade in 1869, but his account of the journey is of limited historical value: Reade, Winwood, The African Sketch-Book (2 vols.: London, 1873), 1:349–510.Google Scholar

71. Binger, L. G., Du Niger au golfe de Guinée (2 vols.: Paris, 1892), 2: 391–92.Google Scholar

72. Arcin, , Histoire, 70.Google Scholar Frobenius' informants explicitly located Sumanguru “towards the south” and Susu west of Niani-Ba, probably on river Sankarani: Frobenius, , Dichten, 329, 332, 336, 338.Google Scholar

73. Bulman, S., “The Buffalo-Woman Tale: Political Imperatives and Narrative Constraints in the Sunjata Epic” in Barber, and Farias, , Discourse, 171, 181Google Scholar

74. Frobenius, , Dichten, 340Google Scholar; Monteil, , “Empires,” 298, 345Google Scholar

75. Bazin, , “Princes désarmés,” 389, 389n46Google Scholar

76. From Sankaran: Quiquandon, , “Histoire,” 307Google Scholar; Innes, , Sunjata, 41, 145, 179Google Scholar; Zeltner, , Contes, 1Google Scholar; Delafosse, , Haut-Sénégal-Niger, 2: 168Google Scholar (king of Sankaran is Sunjata's “uncle”). From Do: Frobenius, , Dichten, 305Google Scholar; Monteil, , “Empires,” 352Google Scholar; Condé, , “Sociétés traditionnelles,” 51.Google Scholar From Sankaran, capital of Do: Cisse, /Kamissoko, , Grande geste, 45, 57.Google Scholar From Do, capital of Sankaran: Frobenius, , Dichten, 331, 338.Google Scholar From Sankaran/Do (interchangeable): Johnson, , Son-Jara, ll 327, 338, 799.Google Scholar From Do equaling Sankaran: Camara, Maître, 37n3, 88; cf. 57. Cf. Bulman, “Buffalo-Woman Tale,” 188nn7-8.

77. Camara, Maître, 37n3.

78. Niane, , Sundiata, 6.Google Scholar

79. Monteil, , “Empires,” 356Google Scholar; Niane, Sundiata, 87n13; Humblot, “Nom,” 528, 529.

80. al-Mukhtar, Ibn, Tarikh el-Fettach, ed. and trans. Houdas, O. and Delafosse, M. (Paris, 1913), 59.Google Scholar Kankan Musa was the legendary emperor and pilgrim Mansa Musa. This figure proliferates, under many different names, in oral traditions of the Western Sudan and beyond.

81. Delafosse, , Haut-Sénégal-Niger, 2: 168–69Google Scholar; Suso, Bamba, Kanute, Banna, and Kanute, Demba in Innes, , Sunjata, 65, 67, 79, 191, 309Google Scholar; Zeltner, , Contes, 31, 42Google Scholar

82. Quiquandon, , Histoire, 310Google Scholar; Frobenius, , Dichten, 330Google Scholar; Innes, , Sunjata, 304Google Scholar

83. A cock's spur was Sumanguru's totem (Sidibé, , “Soundiata Keita,” 44, 45Google Scholar; Niane, , Sundiata, 58Google Scholar). G.D. Pickett, translator of Niane's Sundiata, explains that, by being touched by an arrow with his totem, Sumanguru lost the support of his ancestors (ibid., 93-94n62). Cf. Doumbia, , “Etude,” 347–48.Google Scholar My interpretation that the creature was the spirit of Sumanguru's lineage creates one problem: why would a Konte, such as Ganda, eliminate the ancestral spirit of another Konte, Sumanguru?

84. Frobenius, , Dichten, 329Google Scholar

85. Ibid., 327; cf. Kanute, Banna in Innes, , Sunjata, 179, 191Google Scholar, 195/197: Sankaran Daminya Konte, ancestor of Sunjata's mother and of Sankaran Madiba Konte (or “Ganda”). Delafosse has Danguina Konnté as the king of Sankaran, who is defeated by Sunjata and then joins him (Haut-Sénégal-Niger, 2:178Google Scholar). “Naminya Konde” in Johnson, , Son-Jara, 1. 327Google Scholar, seems to be the same person.

86. Frobenius, , Dichten, 341.Google Scholar Tabung appears to be the “Tabon” of other versions (see below). The clan names of the “hunters” probably refer to royal lineages.

87. For example: Niane, , Sundiata, 55Google Scholar; Delafosse, , Haut-Sénégal-Niger, 2:178.Google Scholar Some authors identify Tabon with Tabu near Sibi, a Kamara territory, and with ironworking and the Jalonke—for example, Person, Yves, “Tradition orale et chronologie,” Cahiers d'études africaines 2 (1962), 465n6CrossRefGoogle Scholar; Niane, , Sundiata, 49.Google Scholar It should be noted, though, that Frobenius' informants in the Siguiri area did not know the location of “Tabung” (Frobenius, , Dichten, 341Google Scholar). A vassal of Tabon was the “Djallonké chief of Neguéboria,” site of a battle between Sunjata and Sumanguru (Niane, , Sundiata, 53, 51Google Scholar).

88. Frobenius, , Dichten, 304Google Scholar; Johnson, , Son-Jara, 1. 799.Google Scholar The number is not to be taken literally: twelve is a magical number often used in oral tradition.

89. Innes, , Sunjata, 271–75Google Scholar

90. Sidibé, , “Soundiata,” 45Google Scholar: Faganda or Fakoli was a leader of the emperor of Susu.

91. An extensive study of the Fakoli figure has recently been presented by Conrad, , “Searching for History.” Pp. 166–74Google Scholar deal with Fakoli's dilemma of loyalty. Relevant excerpts from the Kouroussa version are on p. 170.

92. For example, in Innes, , Sunjata, 276Google Scholar; cf. de Moraes Farias, P.F., “Pilgrimages to ‘Pagan’ Mecca in Mandenka Stories of Origin Reported from Mali and Guinea-Conakry” in Barber, and Farias, , Discourse, 160-61, 165–66.Google Scholar

93. Ly-Tall, Madina, L'Empire du Mali (Dakar, 1977), 194Google Scholar: king Sama of Kangaba moving to Norasoba; Camara, Maître, 38: Samo Condé, king of Do; Johnson, , Son-Jara, 1.Google Scholar 328: Sama Sine, ancestor of tlie Sankaran Konte, ll. 1814, 2586, 2926: (king) Sama, ancestor or son of Sumanguru.

94. Humblot, , “Nom,” 525.Google Scholar

95. Ibid., 528n2 (and bula having same totem as Konte/Sumanguru); Cissé and Kamisoko, Grande geste, 175n38 (bula as hunters/warriors and blacksmiths); Monteil, , “Empires,” 346Google Scholar (bula a slave militia)

96. In Bajar and Kabu, for example, the majority of blacksmiths are Kamara: Simmons, W.S., Eyes of the Night. Witchcraft Among a Senegalese People (Boston, 1971), 80Google Scholar; Carreira, António, Mandingas da Guiné Portuguesa (Bissau, 1947), 158.Google Scholar

97. Tamari, Tal, “The Development of Caste Systems in West Africa,” JAH 32 (1991), 238.CrossRefGoogle Scholar

98. Person, Yves, “Nyaani Mansa Mamudu et la fin de l'empire du Mali” in Sol, parole, écrit, 2: 631.Google Scholar And Arcin noted that, despite tlie vicinity of Niagasola and Kangaba (i.e., the Manding-Keita), the Keitas of both Jumas “ne faisaient plus corps avec eux:” Arcin, André, La Guinée française (Paris, 1907), 297.Google Scholar

99. Person, , “Nyaani Mansa Mamudu,” 2:633.Google Scholar Niani, a presumptive capital of Mali, is located in Juma-Wanya, north of its territorial capital of Sidirila. It appears odd that the Malian emperor, member of the Manding Keita lineage, should have resided in tlie territory of a different Keita lineage.

100. Ibid., 2:631.

101. Although tlie Amana Keita see themselves as distinct from tlie Manding Keita, they nevertheless adopted an origin in Kita, one mythical cradle of the latter. Yves Person perceived this inconsistency to be an “ennui” (“Nyaani Mansa Mamudu,” 631). I believe that it resulted from the desire of tlie Amana Keita to accommodate their own distinct tradition to the prestigious Malian tradition.

102. Jackson, Michael, “The Structure and Significance of Kuranko Clanship,” Africa 44 (1974), 403.CrossRefGoogle Scholar

103. Humblot, , “Nom,” 539.Google Scholar

104. Innes, , Sunjata, 83Google Scholar; Samaké, Maximin, “Kafo et pouvoir lignager chez les Banmana: l'hégémonie gonkòròbi dans le Cendugu,” Cahiers d'études africaines 28 (1988), 335CrossRefGoogle Scholar; Fyle, C. Magbaily, Almamy Suluku of Sierra Leone, c. 1820-1906 (London, 1979), 6Google Scholar; idem., Oral Traditions of Sierra Leone (Niamey, 1979), 26.

105. Person, , Samori, 1: 160, 478n51.Google Scholar

106. Arcin, , Guinée, 58, 310Google Scholar; Person, , “Tradition orale,” 470 (Baleya).Google Scholar

107. Es-Sa'di, , Tarikh es-Soudan, 19–20.Google Scholar

108. Soura and Zara mentioned in ibid., 127, 274. Identification with Jara/Kingi also by Levtzion, , Ancient Ghana and Mali, 98.Google Scholar

109. Es-Sa'di, , Tarikh es-Soudan, 155.Google Scholar

110. Ibid., 278.

111. Delafosse, , Haut-Sénégal-Niger, 2: 208Google Scholar; also Person, , Samori, 1: 87n139.Google Scholar

112. Alvares, Manuel, An Interim Translation of Manuel Alvares S.J., Etiópia Menor e Descripção Geogràfica da Província da Serra Leoa (c.1615), translated and issued by Hair, P.E.H. (Liverpool, 1990), f. 76.Google Scholar

113. Wilks, Ivor, Levtzion, Nehemia, and Haight, Bruce, Chronicles From Gonja. A Tradition of West African Muslim Historiography (Cambridge, 1986), 13, 21–22.Google Scholar

114. al-Mukhtar, Ibn, Tarikh el-Fettach, 68.Google Scholar Bighu was named Bitu (for the identity see Wilks, et al., Chronicles, 7–8Google Scholar). Sankaran was not named in the Tarikh, and only rarely in the Tarikh as-Sudan, because the authors were concerned with the history of Songhay, and the range of its expansion in the Western Sudan included Jara and Futa Toro in the northwest and the eastern provinces of Mali in its west, but not the southwest, where Sankaran was located.

115. Delafosse, , Haut-Sénégal-Niger, 2: 168Google Scholar; cf. Monteil, , “Empires,” 439n1.Google Scholar In one version the Konte ruler of Sankaran volunteers to lead a campaign against Jolof, as did Fakoli and Tiramakan, Sunjata's two ‘standard’ generals (Frobenius, , Dichten, 329Google Scholar). In Delafosse's version a Soninke from Wagadu is Sunjata's lieutenant during a campaign in the north (Delafosse, , Haut-Senegal-Niger, 178–79Google Scholar): an oral vestige of the Faran-Sourâ's role as military governor and army leader in the north?

116. Ganda: Innes, , Sunjata, 65, 67, 79, 191Google Scholar; Kanoute, Dembo, Histoire, 31Google Scholar; Johnson, , Son-Jara, 198Google Scholar; gana, kanda: Delafosse, , Langue mandingue, 2:241.Google Scholar

117. Zeltner, , Contes, 30, 31.Google Scholar In this source the clan name Kanote is identical with Konte because a second informant switches between Kanote and Konté: “Sangaran Madiba Kanoté” and “Sangaran Madiba Konté” (ibid., 41, 42).

118. Condé, , Sociétés traditionnelles, 99Google Scholar; Niane, Sundiata, 87n13.

119. In a Kuranko tradition god made the Mansare (Keita) chiefs, the Sise marabouts, the Konde warriors, and the Mara were given wealth (Jackson, , “Prevented Successions,” 401Google Scholar). See also Doumbia, , “Etude,” 339Google Scholar: Kante originally warriors.

120. Quiquandon, , “Histoire,” 316.Google Scholar

121. Person, , Samori, 36.Google Scholar

122. Frobenius, , Dichten, 332, 336, 338.Google Scholar

123. Corpus, 82.

124. Ibid., 107-08.

125. In Hunwick, , “Géographie”, 420, 426.Google Scholar

126. Corpus, 82.

127. Hunwick, , “Géographie,” 419.Google Scholar

128. Location suggested by ibid., 421.

129. Jahaba, “great Jaha,” was probably the “Jaga” of ca. 1500, (erroneously) termed Mali's capital: Fernandes, Valentim, Description de la côte occidentale d'Afrique (Sènégal au Cape de Monte, Archipels) par Valentim Fernandes (1506-1510), ed. and trans. Monod, T., da Mota, A. Teixeira, and Mauny, R. (Bissau, 1951), 37.Google Scholar

130. This assumption need not contradict Susan Keech McIntosh, who located Wangara in the Niger Inland Delta in her “A Reconsideration of Wangara/Palolus, Island of Gold,” JAH 22 (1981), 145–58.CrossRefGoogle Scholar The name may have been attached to more than one entrepot of the gold trade.

131. Levtzion, , Ancient Ghana and Mali, 54.Google Scholar

132. Al-Bakri, in Corpus, 81.Google Scholar

133. Corpus, 185.

134. In Guerreiro, Fernão, Relação anual das coisas que fizeram os Padres da Companhia de Jesus, ed. Viegas, A. (3 vols.: Coimbra, 1930–1952), 3:240–50.Google Scholar

135. Welmers, William, “Niger-Congo-Mande,” Current Trends in Linguistics 7 (1972), 116Google Scholar; cf. Houis, Maurice, “Que sont les soso?” Etudes guinéennes 6 (1950), 78–79.Google Scholar

136. de Sandoval, Alonzo, Naturaleza, policia sagrada i profana (Seville, 1627)Google Scholar, 40^r: “los Zozoes, que es la casta de los Mandingas corrupta en Zozo.” Unlike other authors, Sandoval had never visited Africa. This may have led Hair to disbelieve that the sentence referred to a linguistic link, “An Interim and Makeshift Edition of André Alvares de Almada's ‘Brief Treatise on the River of Guinea’ (c. 1594),” edited and issued by P.E.H. Hair (Liverpool, 1984/86), part II notes 14/3. But Sandoval had information not provided by other authors and he was particularly well-informed on languages because he catechized slave arrivals in the Caribbean port of Cartagena. For this enterprise he needed African interpreters and some knowledge of the relationships of African languages.

137. As did Houis, Maurice, Etude descriptive de la langue Susu (Dakar, 1963), 2.Google Scholar

138. Pereira, Duarte Pacheco, Esmeraldo de situ orbis par Duarte Pacheco Pereira, ed. and trans. Mauny, R. (Bissau, 1956), 74, 76Google Scholar; de Almada, André Alvares, Tratado breve dos Rios de Guiné do Cabo Verde, ed. Bràsio, António (Lisbon, 1964), 116Google Scholar; Coelho, Francisco de Lemos, Duas descrições seiscentistas da Guiné, ed. Peres, Damião (Lisbon, 1953), 208Google Scholar

139. A French explorer reported Jalonke populations in the area: Caillié, , Journal, 1:260, 262.Google Scholar

140. Sandoval, , Naturaleza, 64r.Google Scholar Sape was a term including different ethnic groups south of the Nunez.

141. Gallieni, , Deux Campagnes, 556-57, 582.Google Scholar

142. Binger, , Niger, 2: 392.Google Scholar

143. Delafosse, , Haut-Sénégal-Niger, 1:127.Google Scholar

144. As already noted by Gallieni, , Deux Campagnes, 540.Google Scholar

145. Its derivation from a toponym “Jalo” led Hair to assume—as I believe, erroneously—that Jalonke referred “to various interior Mande peoples living in the general vicinity of the Futa Jalon” (Hair's annotation of Alvares de Almada, “Interim and Makeshift Edition,” part II, 15/6).

146. Pereira, Pacheco, Esmeraldo, 76.Google Scholar

147. Cortes, Vicenta, La esclavitud en Valencia durante el reinado de los Reyes Católicos (1479-1516) (Valencia, 1964), 280Google Scholar; Bühnen, Stephan, “Ethnic Origins of Peruvian Slaves (1548-1650): Figures for Upper Guinea,” Paideuma 39 (1993), 78–79.Google Scholar It remains to be investigated whether the ethnonyms “Sensones” and “Sonhos” are cognate. They were employed for a Mandinga subgroup along the Gambia and for a group ruled by the king of Bena respectively. Sandoval, , Naturaleza, 48vGoogle Scholar; Bràsio, António, Monumenta Missionaria Africana. Africa ocidental (1600-22) (15 vols.: Lisbon, 1952–1987), 4: 169.Google Scholar

148. See, for example, Arcin, , Guinée, 207Google Scholar; Arcin, , Histoire, 72Google Scholar; Kup, A.P., Sierra Leone: A Concise History (Newton Abbot, 1975), 24Google Scholar; and an oral tradition of Solima in Fyle, , Oral Traditions, 49.Google Scholar

149. Among others, Arcin, , Histoire, 47Google Scholar; Rodney, Walter, “Jihad and Social Revolution in Futa Djalon in die Eighteenth Century,” Journal of the Nigerian Historical Society (1968), 270.Google Scholar

150. See, for example, Amselle, Jean-Loup, “Ethnies et espaces: pour une anthropologie topologique” in Amselle, J.-L. and M'bokolo, Elikia, eds., Au coeur de l'ethnie: Ethnies, tribalisme et etat en Afrique (Paris, 1985), 23.Google Scholar Amselle tends to explain social phenomena, one of which is ethnicity, not as basically emanating from internal processes, but from relations between societies: Segmentary societies result from pressure exerted by states (ibid., 29) and states from conquest and trade (ibid., 27). This view remains attached to the old notions of migration and of ‘state building’ by warriors among conquered non-centralized populations. Recently George Brooks advanced a hypothesis of population movements induced by long-term climatic change and leading to the diffusion of “Mande-speaking groups,” in his “Ecological Perspectives on Mande Population Movements, Commercial Networks, and Settlement Patterns from the Atlantic Wet Phase (ca. 5500-2500 B.C.) to the Present,” HA 16 (1989), 23–40.Google Scholar The use of climatic data for historical reconstruction promises to be of practical value, but I doubt that the emphasis on migration helps in understanding social, political, and technological change.

151. Willis, Justin, “The Makings of a Tribe: Bondei Identities and Histories,” JAH 33 (1992), 193.CrossRefGoogle Scholar

152. Smith, Anthony D., The Ethnic Origins of Nations (Oxford, 1986), 3, 32.Google Scholar

153. This combination of ethnic features had not spread to all populations ruled by Mali during its apogee. Thus the Jolof/Wolof retained their identity, perhaps because its ‘inferiority’ to the Mandinga culture was not perceived as being marked. Yet, they adopted cultural elements deemed Mandinga, among them a multitude of loanwords.

154. Smith's, Anthony term: Ethnic Origins, 43.Google Scholar

155. Suret-Canale, Jean, La République de Guinée (Paris, 1970), 21 (map)Google Scholar; Houis, Etude descriptive; Arcin, , Histoire, 150–59Google Scholar; Delafosse, , Haut-Sénégal-Niger, 1: 151-52, 163–65Google Scholar; Atlas National du Sénégal, ed. Van Chi, R. (Paris, 1977), 67Google Scholar; Person, , Samori, 1:78n22.Google Scholar

156. In 1827 Caillié's African informants spoke of Jalonke as “the ancient possessors of the country of Fouta-Dhialon, which was conquered long ago by the Foulahs” (Caillié, , Journal, 1: 277)Google Scholar; cf. ibid., 363, and Mollien, Gaspard, Travels in the Interior of Africa to the Sources of the Senegal and Gambia (London, 1820), 295.Google Scholar

157. Gallieni, , Deux Campagnes, 490, 496.Google Scholar

158. Pereira, Pacheco, Esmeraldo, 74/76Google Scholar; de Almada, Alvares, Tratado, 116–17Google Scholar; Donelha, André, Descrição da Serra Leoa e dos Rios de Guiné do Cabo Verde (1625) por André Donelha, ed. and ann. da Mota, A. Teixeira, trans, and ann. Hair, P.E.H. (Lisbon, 1977), 98, 162Google Scholar; Coelho, Francisco de Lemos, Duas descriçōes seiscentistas da Guiné de Francisco de Lemos Coelho, ed. Peres, Damiāo (Lisbon, 1953), 59, 208.Google Scholar

159. Mollien, , Travels, 221, 229, 300, 301Google Scholar; Caillié, , Journal, 1: 260-64, 363, 368, 419.Google Scholar

160. Hecquard, Hyacinte, Voyage sur la côte et dans l'interieur de l'Afrique occidentale (Paris, 1855), 365.Google Scholar

161. Suret-Canale, , République, 31Google Scholar; Gallieni, , Deux Campagnes, 466, 488-89, 500, 582Google Scholar; Person, , Samori, 1: 50.Google Scholar

162. Delafosse, , Haut-Sénégal-Niger, 1: 296–97Google Scholar; 2: 361.

163. Arcin, , Histoire, 72Google Scholar; Kup, , Sierra Leone, 44.Google Scholar

164. Binger, , Niger, 2: 302, 391Google Scholar; Arcin, Guinée, map; Person, , Samori, 1: 73Google Scholar

165. Barreira, Baltasar in Guerreiro, , Relação, 3: 244Google Scholar; Alvares, , Etiópia Menor, f. 133v.Google Scholar

166. Innes, , Sunjata, 81.Google Scholar

167. Conrad, , “Searching for History,” 167, 174.Google Scholar

168. Alvares, , Etiópia f. 18vGoogle Scholar, wrote of “Suzes” neighboring Kasa, probably in Birasu. Mandinga-Soninke were first mentioned as “Soniquies, who are commonly called Mandingas” for the River Gambia and the upper Casamance, by Sandoval, , Naturaleza, 5v, 6, 38v.Google Scholar He also wrote of “Mandingas, principalmente Soniquees y Sensones” on both banks of the Gambia (f. 48v). Possibly the latter term is a variant of Soso/Sose. The wording implies a distinction between the Mandingas and the majority Soniquees/Sensones subgroup.

169. Mollien, , Travels, 322Google Scholar: Geba located “in the country of the Saussais Mandingos” and “[m]ost of these Mandingos [of Kabu] are Pagans.” “Soninke:” Coelho, Lemos, Duas descriçōes, 157Google Scholar (Islamic influence of Mandingas on the Soninkes); de Andrade, Bernardino António Alvares, Planta de praça de Bissau e suas adjacentas, ed. Peres, Damiāo (Lisbon, 1952), 56Google Scholar; Park, Mungo, Travels in the Interior Districts of Africa (London, 1792), 35.Google Scholar It should be noted, though, that “Soninke” for pagan Mandinga was also employed as far south as Konya.

170. For a similar case of ethnic change, to a Mandinga identity, see Bühnen, , “Place Names as an Historical Source: An Introduction with Examples From Southern Senegambia and Germany,” HA 19 (1992), 67–69, 85–86.Google Scholar The rapidity of such processes is illustrated by the example of the population of Gaul, which within some few generations following the Roman conquest switched from a Celtic to a Roman identity and language. Roman soldiers and traders were the agents of change.

171. Houis, Maurice, “Quelques données de toponynue ouest-africaine,” BIFAN 20B (1958), 571Google Scholar; Bühnen, , “Place Names,” 79Google Scholar (where I erroneously attributed the suffix to Mandinka).

172. Gomes, , Première découverte, 37Google Scholar (f. 276^v). According to the editors' annotation, “Quioquum” may have been Kukia, capital of the Songhay empire.

173. The tradition mentions an attack at “Oré-Bugu” of Muslim Fula on a caravan of pagan Fula and Jalonke returning from Bundu with merchandise: Moreira, José Mendes, Fulas do Gabú (Bissau, 1948), 252–53.Google Scholar “Horé-Bougou” is north of Labé on a map in Suret-Canale, Jean, “The Western Atlantic Coast, 1600-1800” in History of West Africa, I, ed. Ajahi, J.F.A. and Crowder, M. (London, 1971), 424.Google Scholar

174. The editors of the report already suggested an identity of Gelu with the Futa Jalon: Gomes, Première découverte, 65nn57-58. The different spellings of Geley/Gelu, and the fact that Portuguese ships did not reach Sierra Leone before 1460, suggest that the second passage (“Gelu,” location near Sierra Leone) was inserted into Gomes' report after this date.

175. Pereira, Pacheco, Esmeraldo, 123/125Google Scholar; cf. Fage, J. A., “A Commentary on Duarte Pacheco Pereira's Account of the Lower Guinea Coastlands in his Esmeraldo de Situ Orbis and on Some Other Early Accounts,” HA 7 (1980), 53–54.Google Scholar P.E.H. Hair was sceptical whether “genuine Susu came as far east:” see his “Ethnolinguistic Continuity on the Guinea Coast,” JAH 8 (1967), 260.Google Scholar

176. Pereira, Pacheco, Esmeraldo, 74, 76, 84.Google Scholar

177. Magalhāes-Goudinho, Vitorino, L'économie de l'empire portugais aux XVe et XVIe siècles (Paris, 1969), 209.Google Scholar

178. The gold bought at Sierra Leone was known as “the finest gold” in Guinea (Fernandes, , Description, 96Google Scholar). Likewise, gold bought south of Cape Mount was reported to be “very fine,” of “23 carats” (Pereira, Pacheco, Esmeraldo, 100Google Scholar). This latter gold must have originated from the mines near the upper St. Paul.

179. Fernandes, , Description, 76, 48.Google Scholar

180. One major center of ancient (and recent) iron extraction is the north of modern Sierra Leone in the Loko and Limba areas (Donelha, , Descrição, 104Google Scholar).

181. Caillié, , Journal, 1:469.Google Scholar

182. de Almada, Alvares, Tratado, 116–17Google Scholar (the “Putazes” were Susu: Alvares, , Etiópia, f. 133)Google Scholar, 125; Donelha, , Descrição, 98, 162.Google Scholar

183. Coelho, Lemos, Duas descrições, 208Google Scholar; de Anguiano, Mateo, Misiones Capuchinas en Africa, ed. de Carrocera, B. (2 vols.: Madrid, 1950–1957), 2: 132.Google Scholar

184. de Almada, Alvares, Tratado, 127 (Conchos)Google Scholar; Donelha, , Descrição, 96, 118Google Scholar, see also P.E.H. Hair's note, 237n113; Francisco d'Andrade 1582, Bartolomeu André, 1606 and Baltasar Barreira, 1606 in Bràsio, , Monumenta Missionaria, 3:106Google Scholar; 4:117 (all gold of Sierra Leone from Concho via Bena), 171 (iron mines, Susu iron superior to other provenances); Alvares, , Etiópia, f. 133Google Scholar (gold coming “by way of the Fulos and Mandingas,” iron, salt).

185. Alvares, , Etiópia, f. 76, note a.Google Scholar

186. Magalhães-Goudinho, , Economie, 538–39Google Scholar; Mauny, Raymond, Tableau géographique de l'ouest africain au moyen âge (Dakar, 1961), 249.Google Scholar

187. Alvares, , Etiópia, f. 133, 133vGoogle Scholar; Laing, , Travels, 81, 372Google Scholar; Caillié, , Journal, 1: 346.Google Scholar

188. Alvares, , Etiópia, f. 133v.Google Scholar

189. Gallieni, , Deux Campagnes, 615Google Scholar; Famechon, , Notice, 177Google Scholar; Arcin, , Guinée, 100, 102, 104Google Scholar; Binet, J., “Marchés en pays Soussou,” Cahiers d'études africaines 3 (1962), 110.CrossRefGoogle Scholar

190. Alvares, , Etiópia, 134v.Google Scholar

191. Bradbury, R. E., “The Historical Uses of Comparative Etlinography with Special Reference to Benin and the Yoruba” in The Historian in Tropical Africa, ed. Vansina, J., Mauny, R., and Thomas, L.-V. (London, 1964), 154, 159Google Scholar; Goody, Jack, Technology, Tradition, and the State in Africa (London, 1971), 53.Google Scholar

192. Alvares, Etiópia, f. 133v-34 (additions in brackets by the translator P.E.H. Hair).

193. Thus, these primary polities had not been “mis en place par une instance étatique, l'ancien Mali” (Bazin in his “Princes,” 398). The general rule of persistence of one royal lineage within a territory has its exceptions, but a change of a royal lineage is often retained in the collective memory because sacral ties to the earth continue to be held by the deposed kin group, whose elder officiates at community rituals addressing spirits of the earth.

194. Compare the equation of “duke” and silatigi, on the Gambia, in Donelha, , Descriçāo, 148Google Scholar, and note 255 by P. E. H. Hair.

195. Gallieni, , Deux Campagnes, 305Google Scholar (Sangala: majority Kamara), 539 (Tamiso), 556 (Baleya), 599 (Bure); Person, Samori, 87n142 (Bure, Baleya, Firiya), 129n83 (Bure).

196. Arcin, , Guinée, 311Google Scholar; Kup, , Sierra Leone, 42Google Scholar; Fyle, , Almamy Suluku, 6.Google Scholar

197. According to one tradition, a Keita Jalonke king, the “Manga Labe,” ruled during the jihad: (Suret-Canale, , Guinée, 32Google Scholar). Susu/Jalonke manga, mange, “king.”

198. Person, , “Nyaani Mansa Mamudu,” 631 (Amana, Juma)Google Scholar; Person, Samori, 125n39 (Huré), 324 (Baleya, Tumania); Gallieni, , Deux Campagnes, 467 (Meretambaya).Google Scholar

199. Magalhães-Goudinho, , Economie, 209.Google Scholar

200. Donelha, , Descrição, 120 (my translation differs from Hair's on p. 121).Google Scholar

201. Alvares, Etiópia, f. 133v; Barreira, Baltasar in Guerreiro, , Relação, 3:244.Google Scholar

202. Coelho, Lemos, Duas descrições, 88Google Scholar (“Great Concho” little known, extensive territory), 235 (Great Limba and Great Concho “according to the blacks are very extensive kingdoms”); Anguiano, , Misiones, 132 (Conchos over Susus and Fulas).Google Scholar

203. Moreira, Mendes, Fulas, 265.Google Scholar The source is a written chronicle of the Fula chief of Madina do Boé in the east of what is now Guinea-Bissau.

204. de Almada, Alvares, Tratado, 127Google Scholar; Donelha, , Duas descrições, 96Google Scholar; Anguiano, , Misiones, 2: 132 (traders, different in color and hairstyle).Google Scholar

205. de Almada, Alvares, Tratado, 127Google Scholar; Coelho, Lemos, Duas descriçōes, 65.Google Scholar

206. Bràsio, , Monumenta, 4: 169.Google Scholar

207. Es-Sadi, , Tarikh es-Soudan, 279Google Scholar (on the Jenne affair). According to Dapper, kingdoms along the Gambia had become independent “a few years ago:” Dapper, Olfert, Naukeurige Beschrijvinge der Afrikaensche Gewesten (2d ed.: Amsterdam, 1676), 415, 2.Google Scholar His source for the Gambia was probably one Willem Block, who was in the region in 1615 (reference for date in Thilmans, G., “Le Sénégal dans l'oeuvre d'Olfried Dapper,” BIFAN 33B [1971], 545n4bis).Google Scholar The fairly precise dating, “a few years ago,” suggests an abrupt end of Malian supremacy over polities along the Gambia.

208. There is no previous passage dealing with the “Province of the Sousos” in the edition. Perhaps it was part of ff. 44-46v, which are missing. Following the geographical order of Alvares' account from north to south, these probably treated the Beafada in more detail, as well as the Nalu, Baga, and Susu.

209. Alvares, , Etiópia, f. 76.Google Scholar

210. Ibid., f. 133v.

211. Ibid., f. 76. Compare the golden arms of the Malian king (al-'Umari in Corpus, 265) and a golden and a silver lance held by his griot at a ceremony (Ibn Battuta in Corpus, 290), resembling the employment of such sacral regalia in southern Senegambia by the king's head slave or his marabout. The supreme rank of golden regalia is reflected in the imagery of a tradition narrated by Kanoute, , Histoire, 57Google Scholar: Sunjata received the defeated King of Jolof's cap, slippers, arrow, and spear, all made of gold, illustrating Jolofs new state of vassalage to Mali. Golden regalia ranked above regalia made of, in declining order, silver, brass, iron, and wood. Among the insignia of the (Beafada) King of Guinala, south of Rio Geba, were an iron bow and iron arrows (Donelha, , Descrição, 176Google Scholar). The King of Kabu, intermediate between the emperor of Mali and the King of Guinala, had a silver lance: Niane, D. T., Histoire des Mandingues de l'Ouest: Le royaume du Gabou (Paris, 1989), 68.Google Scholar For the ranking of silver, iron, and wooden staffs see Peter Weil, “The Chono: Symbol and Process in Authority Distribution in Mandinka Political Entities of Senegambia,” paper presented at the Southwestern Anthropological Association Meeting, San Francisco, 1973.

212. Alvares, , Etiópia, f. 76.Google Scholar The hinterland of “Mina” (and the Liberian “Malaguetta Coast”) was also named for the invasion route of the Upper Guinean Mane, whom a tradition had as originating from Congo (Donelha, , Descrição, 106Google Scholar; de Almada, Alvares, Tratado, 133Google Scholar). Cf. Hair's annotation of Alvares de Almada, “Interim and Makeshift Edition,” part II, notes, 16/7, and Rodney, Walter, A History of the Upper Guinea Coast, 1545-1800 (Oxford, 1970), 39–44.Google Scholar

213. Fernandes, , Description, 74Google Scholar; Pereira, Pacheco, Esmeraldo, 74.Google Scholar

214. Barreira, in Guerreiro, , Relação, 3: 243–44.Google Scholar “Bexerin” was northern Senegambian and, as a loanword, Upper Guinean Portuguese for “marabout.”

215. Alvares, , Etiópia, f. 136.Google Scholar In southern Senegambia pagan societies influenced by Islam and Mali had a seven-day week (Mandinga) while societies less, or not, exposed to this influence had six-day weeks (Kasanga, Bainunk, Manjak/Papel, Beafada, Diola). Friday is the holy day of the Muslims, Thursday is associated with the staging of (pagan) masks, and Monday was the day of the king's rituals: Frobenius, , Spielmannsgeschichten, 32Google Scholar (annual iron-smelting ceremony with koma masks on a Thursday); Weil, Peter, “Men's Masking and Ritual in the 19th and 20th Century: Adaptive Processes of the Mandinka of Senegambia,” paper presented at the Annual Meeting of the African Studies Association, Madison, 1986 (Wuli: staging of masks on Thursdays, Monday “king's ritual day”)Google Scholar; Gray, William and Dochard, N., Travels in Western Africa (London, 1825), 302Google Scholar (Kaarta: Monday “his Majesty's drinking day”); Innes, , Sunjata, 56/57, 178/179, 190/191Google Scholar (griots allude to installation of kings on Mondays); Jobson, Richard, The Golden Trade (London, 1623), 156Google Scholar (Gambia: a market held every Monday, an indirect hint to Monday as the king's day).

216. “Putazes” or “Putas” were “virtually Sousos too” and were ruled by the “Farim Puta” (Alvares, , Etiópia, f. 133Google Scholar, cf. ff. 88, 133v). “Farim Caputa…rules over the hinterland of the Bagas” (Donelha, , Descrição, 120Google Scholar). I interpret “Caputa” as “Puta land,” as did Hair in his annotation of Alvares de Almada's Tratado in “Interim and Makeshift Edition,” part II, notes, 14/3. The Puta must have been a section of the Susu living behind the Baga near river Nunez.

217. de Almada, Alvares, Tratado, 125Google Scholar, cf. ibid., 116-17 for the “Putas.” Possibly in the later fifteenth century a Fula army had ravaged southern Senegambia and probably the Futa Jalon. For a discussion of the campaign and further references see Boulègue, Jean, Le Grand Jolof (XIIIe-XVI siècle) (Paris, 1987), 156–60.Google Scholar

218. de Almada, Alvares, Tratado, 142, 144.Google Scholar This may indicate that the Susu did not fight on horseback, at least not south of the Futa Jalon.

219. Coelho, Lemos, Duas descriçōes, 59, 208.Google Scholar

220. Anguiano, , Misiones, 132 (based on information collected before 1687).Google Scholar

221. Walter Rodney wrote a convincing outline of changes in the relations between Fula and Jalonke leading to the jihad in his “Jihad,” 271-77.

222. The Mandinga inhabitants of Kabu were called Jalonke in the 1880s (Gallieni, , Deux Campagnes, 497Google Scholar), probably to emphasize their former heathendom well-known in the region from the jihad of the Kabu Fulas ending in the 1860s. For late Jalonke conversion see Arcin, , Guinée, 511, 520Google Scholar; Suret-Canale, , Guinée, 63.Google Scholar

223. See, for example, Gallieni, , Deux Campagnes, 304-05, 467Google Scholar; Arcin, , Histoire, 105.Google Scholar

224. E.g., ibid. 90-91.

225. Moreira, Mendes, Fulas, 265.Google Scholar

226. An oral tradition has the kusa, a Soninke group of clans, originate from Diongodji south of Nioro-du-Sahel. After their dispersal “c'est Djallon Magha qui régna à Diongodji.” According to a footnote referring to this last line of the edition, “[l]e recit se poursuit une demi-heure encore par la légende de Jallo Maxa, puis par celle de Sujatta” (Meillassoux, , Légende, 133Google Scholar). The Wolof and northern Malinke word maxa means “king.” Hence “Jallo Maxa” sounds like a northern version of “Jalomansa.” This ruler's chronological position between the kusa-Soninke and Sunjata (“puis par celle de Sujatta”) may mean that this part of the tradition deals with Susu, because both Ibn Khaldun and the Sunjata epic place Susu's rule before that of Mali/Sunjata.

227. Amselle, Jean-Loup, Les négociants de la savane: histoire et organisation sociale des Kooroko (Mali) (Paris, 1977), 46-65, 106Google Scholar

228. Doumbia, , “Etude,” 346–47.Google Scholar

229. Laing, , Travels, 81, 351, 357, 372.Google Scholar The Jalonke of Baleya were also mentioned as producing and trading white cloth (Caillié, , Journal, 1:346, 363Google Scholar).

230. Laing, , Travels, 195, 279, 280, 344, 371.Google Scholar

231. Ibid., 356-57. Evidence for the competition for trade routes between groups of traders and for rulers prescribing trade routes in ibid., passim (Kuranko, Solimana), and in Arcin, , Guinée, 104 (Susu west of Futa Jalon).Google Scholar

232 Caillié, , Journal, 1:335.Google Scholar

233 Zweifel, J. and Moustier, M., Voyage aux sources du Niger (Marseille, 1880), 30, 34, 59, 82, 92, 99, 153.Google Scholar

234 Ibid., 151.

235. For an attempt to date the conquest of Ghana by Susu see Levtzion, Nehemia, “Ancient Ghana: A Reassessment of Some Arabic Sources” in Sol parole, ecrit, 436–37.Google Scholar